LARES: Latent Reasoning for Sequential Recommendation

2025-05-22

AI Summary

Existing sequential recommendation methods suffer from inadequate modeling of dynamic user interests and rely on non-inferential paradigms, limiting both representation expressiveness and computational efficiency. Method: This paper proposes a lightweight framework based on deep recursive implicit reasoning. It introduces (1) a parameter-free-growth implicit reasoning paradigm that progressively refines tokens layer-by-layer to enhance computational density and representation capacity; and (2) a two-stage training strategy comprising trajectory-level and step-level dual-alignment self-supervised pre-training (SPT) followed by reinforcement learning–based post-training (RPT). Contribution/Results: The framework achieves significant improvements over state-of-the-art methods across multiple real-world benchmarks. Moreover, it is modular and plug-and-play—directly boosting the performance of existing advanced models without architectural modification.

Technology Category

• Reasoning under Uncertainty: Sequential Decision Making
• Search and Optimization: Learning to Search
• Machine Learning: Learning Preferences or Rankings

Application Category

• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Sequential recommender systems have become increasingly important in real-world applications that model user behavior sequences to predict their preferences. However, existing sequential recommendation methods predominantly rely on non-reasoning paradigms, which may limit the model's computational capacity and result in suboptimal recommendation performance. To address these limitations, we present LARES, a novel and scalable LAtent REasoning framework for Sequential recommendation that enhances model's representation capabilities through increasing the computation density of parameters by depth-recurrent latent reasoning. Our proposed approach employs a recurrent architecture that allows flexible expansion of reasoning depth without increasing parameter complexity, thereby effectively capturing dynamic and intricate user interest patterns. A key difference of LARES lies in refining all input tokens at each implicit reasoning step to improve the computation utilization. To fully unlock the model's reasoning potential, we design a two-phase training strategy: (1) Self-supervised pre-training (SPT) with dual alignment objectives; (2) Reinforcement post-training (RPT). During the first phase, we introduce trajectory-level alignment and step-level alignment objectives, which enable the model to learn recommendation-oriented latent reasoning patterns without requiring supplementary annotated data. The subsequent phase utilizes reinforcement learning (RL) to harness the model's exploratory ability, further refining its reasoning capabilities. Comprehensive experiments on real-world benchmarks demonstrate our framework's superior performance. Notably, LARES exhibits seamless compatibility with existing advanced models, further improving their recommendation performance.

Problem

• Enhancing sequential recommendation via latent reasoning to improve performance
• Addressing non-reasoning paradigms limiting computational capacity in recommendation systems
• Developing scalable framework with depth-recurrent reasoning for dynamic user interests

Innovation

• Depth-recurrent latent reasoning enhances computation density
• Two-phase training with self-supervision and reinforcement learning
• Refines input tokens at each implicit reasoning step